Wheat germplasm resources are extremely important for breeders. One important wheat-producing area of the world is located between 30° and 45°N latitude, which is rich in wheat germplasm resources. A prerequisite for making full use of these germplasm resources is to assess their genetic diversity [31]. We used high-throughput sequencing technology to discover a large number of SNPs for genotyping hexaploid wheat derived from diverse provenances in this study. Rufo et al. (2019) [32] genotyped a set of 354 Mediterranean wheat accessions using 13,177 SNP markers and observed that the highest number of SNPs were mapped to the B genome and the lowest number to the D genome. Chao et al. (2009) [33] and Berkman et al. (2013) [34] showed that the highest number of SNPs was mapped to the B genome, followed by the A genome and D genome. However, Cavanagh et al. (2013) [35] reported that the majority of SNPs were located in the A genome and the lowest number in the D genome. Zhang et al. (2017) [10] used 271 SSR markers to detect the highest number of alleles in the D genome, the lowest number were in the A genome. The highest average number of alleles was detected in the B genome, followed by the D genome and the A genome, using 17 SSR markers, as reported by Salem et al. (2015) [36]. In the present study, we obtained 24,767 SNPs markers and observed the lowest frequency of SNPs in the D genomes, whereas the B genome contained the highest frequency of polymorphic markers, which is in agreement with the results of previous studies [1, 19, 28, 32, 37, 38, 39–41]. Furthermore, the fewest SNP markers were located on chromosome 4D, whereas the highest number of SNP markers were located on chromosome 3B, as reported by Saintenac et al. (2013) [42] and Alipour et al. (2017) [28]. Eltaher et al. (2018) [19] obtained 25,566 SNPs by GBS for 270 F3:6 Nebraska winter wheat accessions, and observed that the highest number of SNPs were located on chromosome 3B, whereas chromosome 3D carried the lowest number of SNPs. Bhatta et al. (2017) [2] reported that chromosomes 2B and 4D had the highest and lowest numbers of SNPs, respectively. Chromosome 4D had the lowest number of markers and chromosome 1B had the highest number of markers in the study by Sukumaran et al. (2015) [43]. Allen et al. (2017) [44] used 35,143 SNPs and reported that chromosome 2B had the highest number of markers and chromosome 4D had the lowest number of markers. In contrast, the present study showed that chromosome 3B harbors the highest number of SNPs and chromosome 4D has the lowest number.
The PIC contributes to a detailed understanding of the level of polymorphism between genotypes. On the basis of previous reports, the PIC can be divided into three categories: (1) when PIC > 0.5, the marker is considered to be highly polymorphic, (2) when 0.25 < PIC < 0.5, the marker is a moderately informative, and (3) when PIC < 0.25, the marker is a low-information marker [45]. Lopes et al. (2015) [46] observed a PIC value of 0.27 using the 9K SNP array to genotype the WAMI population, and showed that spring wheat contained moderate levels of polymorphism. Novoselović et al. (2016) [47] genotyped a Croatian panel using a set of 1229 Diversity Arrays Technology (DArT) markers and obtained an average PIC value of 0.30 among the populations, which indicated that the accessions from Croatia exhibited moderate polymorphism. Eltaher et al. (2018) [19] analyzed 270 F3:6 Nebraska winter wheat accessions, and observed a PIC value of 0.25, which indicated that the population contained moderate genetic diversity. El-Esawi et al. (2018) [48] used 1052 DArT markers to genotype Australian and Belgian wheat accessions, and obtained PIC values of 0.33 and 0.29, respectively, which demonstrated that Australian and Belgian wheat contain moderate genetic diversity. The present results showed that the mean PIC value (0.27) was in agreement with the above-mentioned studies, which indicated that the 180 accessions contained moderate polymorphism. On the other hand, Hao et al. (2011) [49] gemotyped 250 Chinese wheat accessions with 512 SSR markers and observed a PIC value of 0.650, which demonstrated that Chinese wheat shows high genetic polymorphism. Zhang et al. (2010) [50] analyzed 205 elite wheat accessions in the USA, using 245 SSR markers, and obtained a PIC value of 0.54, which indicated that the accessions showed a high level of polymorphism. Relative to SSR markers, the lower PIC value of the SNP and DArT markers may be explained by their bi-allelic nature and slow mutation rate [51, 52].
In the present study, we obtained meaningful information on genetic diversity indices in each subpopulation. High levels of genetic diversity were represented in Groups 1 and 2, with genetic diversity detected in Group 2. The results of AMOVA showed that a high level of genetic diversity was observed within subpopulations, whereas the variation among subpopulations was extremely low (1%). This result may be caused by breeders selecting for specific traits, such as yield, stripe rust resistance, and herbicide tolerance. However, the low genetic variability among subpopulations is explained by the high gene flow [53]. Wright (1965) [54] showed that when Nm (haploid) values are less than 1, gene exchange among subpopulations is limited. In the present study we observed an extremely high Nm value (28.124), indicating that high gene flow led to low genetic differentiation among subpopulations. The results of this study will not only help breeders to understand the genetic diversity of wheat germplasm on the Eurasian continent between the latitudes of 30° and 45°N, but also provides valuable information for genetic improvement of wheat through inclusion of novel genetic variation from China and certain other countries.
The PCA revealed a degree of broad geographic partitioning of the accessions. A previous study by Winfield et al. (2018) [6] used 32,443 polymorphic markers to genotype 804 hexaploid wheat accessions originating from more than 30 countries around the world, and observed that the majority of accessions from Europe clustered together, separate from the majority of Asian and Middle Eastern accessions. Similarly, in the study of Cavanagh et al. (2013) [35], the European winter wheat population showed the strongest degree of genetic differentiation from the remaining populations. Balfourier et al. (2007) [55] used a set of 38 SSR markers to analyze 3942 accessions originating from 73 countries, and observed that accessions from several Near Eastern and Central Asian areas were grouped in the same subcluster and those from Far Eastern countries clustered together. Strelchenko et al. (2005) [56] analyzed 78 wheat landraces originating from 22 countries and reported that the landraces were separated into European and Asian groups. Chen et al. (2019) [57] reported that West Asian landraces, the majority of European landraces, several South and Central Asian landraces, and the majority of East Asian cultivars clustered together, whereas the majority of East Asian landraces were clustered with several West Asian landraces and the majority of South and Central Asian landraces. Lee et al. (2018) [58] reported that many accessions from Afghanistan, Japan, and Korea were clustered in the same group, while germplasm from China, the Middle East, and Caucasus clustered in a separate group, and an intermediate group largely consisted mainly of accessions from Afghanistan, Japan, and Korea. In the present study, although there was substantial overlap between clusters, the majority of accessions from Europe clustered together, whereas the accessions from Asia and the Middle East were distributed evenly on PC1 (Fig. 4). However, the relationships of three overlapping subgroups was unclear, which raises the possibility of exchanging adapted germplasm. To obtain useful information on the genetic diversity and population structure of the accessions, they were divided into two subgroups on the basis of the population structure analysis (Fig. 5). In the PCA (Fig. 6), genotypes clustered consistent with the subpopulations identified in the STRUCTURE analysis. Moreover, the UPGMA cluster analysis (Additional file 1: Fig S1) was consistent with the results of the STRUCTURE analysis. The majority of European accessions were divided into Group 1, especially accessions from Bulgaria and Portugal (Fig. 8), whereas portions of the Asian and Middle Eastern accessions were distributed in Groups 1 and 2, respectively. The accessions from Turkey, Syria, Georgia, Armenia, Afghanistan, Kyrgyzstan, and Tajikistan showed complex genetic backgrounds, which is not surprising. The area between the Black Sea and the Caspian Sea, and just south of this region (Iraq), is the assumed location of the center of origin of wheat domestication and seems to be a site of population consolidation. Chen et al. (2019) [57] showed that Chinese wheat accessions were mainly derived from European landraces. In the present study, the accessions originating from China tended to cluster with European accessions (Fig. 8).